Predictive modeling of HVOF-sprayed TiC coating: an ANN-based approach for coating properties estimation

This research article presents a comprehensive investigation into the High-Velocity Oxy-Fuel spraying process, focusing on the creation of coatings using SS316 as the base material and Titanium Carbide as the coating powder. The study systematically explores the influence of key process parameters, including oxygen flow rate (O), LPG flow rate (L), and air flow rate (A), on critical coating properties such as coating thickness, porosity, and slurry erosion resistance. To gain insights and predict coating properties accurately, an Artificial Neural Network (ANN)-based regression model is developed. The ANN model is meticulously optimized, with a single hidden layer containing 20 neurons identified as the most effective architecture. The model demonstrates strong performance in fitting training data and accurately predicting coating characteristics. Validation of the ANN model is conducted, revealing close agreement between model predictions and experimental observations. Scanning Electron Microscope images, porosity analysis, and mass loss measurements further corroborate the model's precision in estimating coating properties. The study underscores the utility of data-driven approaches, particularly ANN-based regression models, in materials science research, offering a systematic and reliable means of predicting coating properties without relying on complex physical models.